Artificial intelligence (AI) is revolutionizing the field of medicine, particularly in its ability to predict patient outcomes. Three recent developments highlight the potential of AI in this regard.
One interesting possibility is the use of AI systems, such as ChatGPT and Google's Bard, to collate data from various sources and provide patients with information on doctors and hospitals that offer the highest chances of treatment success for their specific conditions. This could be accomplished by identifying surgeons with the most experience in a particular procedure or hospitals with excellent survival rates for specific diseases. However, it is important to note that the reliability of the underlying data and the context in which it is presented must be considered. Complication rates, for example, may not necessarily reflect the competence of a physician and could be influenced by factors such as the severity of patients' conditions. Nevertheless, the transparency of success rates and complication rates could be a positive development.
Another intriguing advancement is the use of AI algorithms to detect individuals with type 2 diabetes through voice analysis. Researchers have found that subtle changes in voice intensity, amplitude, and pitch variation can indicate the presence of the disease. The accuracy of computer predictions based on these vocal cues was around 86% for men and 89% for women. Although the exact mechanism behind this detection method is not fully understood, it is believed that early diabetes affects the mechanical properties of vocal cords and patients' ability to control their vocal muscles. If further studies confirm these findings, this non-invasive and inexpensive screening method could significantly benefit millions of Americans affected by diabetes.
Additionally, researchers at Vanderbilt University and the University of Missouri-Kansas City have demonstrated that data from CT scans of the chest used to screen for early lung cancers can also predict mortality due to lung cancer, cardiovascular disease, or any cause. This information can help physicians identify patients who would benefit from interventions like physical conditioning or lifestyle modifications, even before the onset of disease.
While AI is not yet reliable for predicting short-term mortality in emergency settings, researchers are exploring the ethical implications of algorithms that could make such predictions. There is recognition among healthcare professionals that AI holds promise in directing appropriate care for high-mortality-risk patients, although concerns arise if its implementation is solely driven by cost-saving motives.
As AI continues to improve its predictive capabilities in healthcare, patients and physicians will encounter new opportunities and challenges in utilizing this knowledge. While it is essential to approach these advancements with caution and consider the limitations and ethical implications, the potential benefits in improving patient outcomes are substantial.